Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Increased activation of frontal native-language areas after second-language learning: A precision fMRI study
Poster Session D, Saturday, September 13, 5:00 - 6:30 pm, Field House
M. Ryan Henderson1, Rebecca Belisle1, Jennifer Minas2, Zhenghan Qi2,3, Amy Finn2,4, John Gabrieli2, Tyler Perrachione1,2; 1Boston University, 2Massachusetts Institute of Technology, 3Northeastern University, 4University of Toronto
Second-language learning causes widespread changes in functional brain activation. However, whether increased responses to a newly learned language are seen in the same brain areas that are functionally selective for the native language remains unknown: Prior work investigating the brain bases of second language acquisition has predominately relied on group averages that obscure variation in local functional organization across individuals. Recent advances in precision neuroimaging enable investigations of the functional organization of the language network in individuals (Fedorenko et al., 2024). Here, we apply a precision neuroimaging approach to the question of second language learning, asking whether areas that are preferentially responsive to a participant’s native language also become responsive to second languages after learning. We recruited two groups of participants who completed different training paradigms: One group (n=41) underwent a four-day laboratory-based training to learn a miniature artificial language (MAL). The MAL differed from English in its lexicon, word order, and inflectional morphology for noun class and verb agreement. The second group (n=30) underwent a month-long classroom-based training to learn Mandarin Chinese. We scanned participants before and after training using a language-comprehension task, in which they matched sentences they heard with pictures. All participants completed this task in English and both foreign languages. Before training, the MAL and Mandarin conditions were meaningless to all participants, and accuracy in both was at chance (vs. 90% in English). After training, participants understood the new language they had learned (80% accuracy), but the untrained language remained meaningless (chance performance). We identified subject-specific fROIs within 13 probabilistic language areas based on a population-level atlas (Lipkin et al., 2022): left and right aSTG, pSTG, IFG pars opercularis, IFG pars triangularis, precentral gyrus; left supplementary motor area and superior frontal gyrus; and right cerebellum. Language-selective fROIs were defined independently in each run of the pre-training data by contrasting activation to English vs. the average of the two unknown languages. We used these fROIs to independently sample activation from each condition during the other pre-training run and both post-training runs. We investigated training-related activation differences between conditions and groups within each fROI using linear mixed-effects models. We found that, after training, core nodes of the language network exhibited increased responses to the learned language, but not the unlearned language. Interestingly, these changes were limited to left frontal areas and cerebellum, but not right frontal homologues or left or right temporal areas. This might suggest an important role for left frontal regions in facilitating the development of high-level comprehension of a second language at early stages of proficiency. Temporal regions may not have shown significant learning effects due to high levels of mean activation within those fROIs for all conditions during both scanning sessions. In further analyses with these data, we will investigate whether there are changes in local activation patterns to trained vs. untrained languages within temporal regions using MVPA. We will also examine whether there are dissociable changes specific to learning a tonal language like Mandarin, or more grammatically-complex languages like the MAL.
Topic Areas: Language Development/Acquisition, Multilingualism